Hyperspectral image compression method based on fractal multi-wavelet

Abstract

The invention provides a hyperspectral image compression method based on fractal multi-wavelet, and the method comprises the steps: grouping to-be-compressed three-dimensional hyperspectral images through a 3D SPIHT image compression system, and carrying out the fractal multi-wavelet transformation of the expanded to-be-compressed three-dimensional hyperspectral images through a constructed fractal multi-wavelet filter matrix; and finally, obtaining a 3D SPIHT hyperspectral image compression code stream based on fractal multi-wavelet by using the quantization coefficient. According to the hyperspectral image compression method based on fractal multi-wavelet, 3D SPIHT based on fractal multi-wavelet transform is adopted, a hierarchical tree set partitioning (SPIHT) coding method is adopted,so the distribution characteristics of fractal multi-wavelet coefficients of the hyperspectral image and the relevance of the wavelet coefficients of the hyperspectral image between different scales can be well utilized; moreover, the relevance of wavelet coefficients of the hyperspectral image in the same scale is considered as much as possible, scalable compression coding of the hyperspectral image is completed, and higher compression coding efficiency is achieved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a hyperspectral image compression method based on fractal multi-wavelets. Background technique [0002] As an emerging remote sensing technology, hyperspectral imagery can provide more information that people need due to the addition of one-dimensional data, and is widely used in agriculture, industry, and military fields, but the rich information also leads to the expansion of data volume , requires special compression methods for effective encoding in order to store and transmit smoothly. Hierarchical Tree Set Partitioning (SPIHT) image compression method indirectly realizes the bit plane sorting of spatial wavelet, effectively utilizes the multi-resolution characteristics of wavelet decomposition, and has the characteristics of embedded code stream. Therefore, SPIHT image compression method has the ability to restore image quality High, easy to output bit rate precise control ...

AI Technical Summary

Problems solved by technology

However, the Haar wavelet and 9 / 7 wavelet are single wavelets, and it is impossible to have the symmetry, short support, orthogonality and high-order vanishing moment properties required to represent hyperspectral images at the same time. Each layer of wavelet transform can only obtain 4 different The characteristics in the direction, the concentration of energy and entropy after wavelet transform is not high, and the wavelet transform needs to extend the boundary of the image to cause boundary distortion, which is not conducive to block processing and parallel acceleration

Images